Search: “perl”

CVE-2026-85485: HTML::FormHandler versions before 0.410002 for Perl render some error messages into HTML without escaping

HTML::FormHandler before 0.410002 for Perl renders error messages into HTML unescaped, enabling cross-site scripting (CVE-2026-85485); fixed in 0.410002.

Robert Rothenberg disclosed CVE-2026-85485 on the oss-security mailing list, affecting HTML::FormHandler versions before 0.410002 for Perl. The module renders some error messages into HTML without escaping, which can enable cross-site scripting in applications using its rendered output. The issue is fixed in version 0.410002.

oss-security·Sep 8, 20:28 UTC · 7d agoVulnerability 2 sourcesCVE-2026-854851

CVE-2026-85484: HTML::FormHandler versions before 0.410002 for Perl render option group labels and radio button labels into HTML without escaping

HTML::FormHandler before 0.410002 for Perl renders option group and radio button labels unescaped into HTML, enabling XSS (CVE-2026-85484).

Robert Rothenberg disclosed CVE-2026-85484 on the oss-security mailing list, affecting HTML::FormHandler versions before 0.410002 for Perl. The module renders option group labels and radio button labels into HTML without escaping, which can enable cross-site scripting in applications using the rendered markup. The issue is addressed in version 0.410002.

oss-security·Sep 8, 20:16 UTC · 7d agoVulnerability 2 sourcesCVE-2026-85484

USN-8675-2: Perl vulnerabilities

Ubuntu issued USN-8675-2 fixing two Perl flaws (CVE-2026-12087, CVE-2026-13221) enabling information disclosure and regex-based security bypass on 26.04 LTS.

Ubuntu released USN-8675-2, extending the fixes from USN-8675-1 to Perl packages on Ubuntu 26.04 LTS. The update addresses CVE-2026-12087, an out-of-bounds heap read in the Socket module when handling short source addresses, which could lead to information disclosure. It also fixes CVE-2026-13221, where regular expressions containing many fixed string alternatives could produce incorrect matches and bypass security restrictions. No exploitation is reported in the notice.

Ubuntu Security Notices·updated Sep 11, 12:03 UTC · 5d ago·first Sep 9, 14:10 UTC · 6d agoAdvisory 13 sourcesCVE-2026-12087CVE-2026-13221

USN-8736-1: Perl vulnerabilities

Ubuntu issued USN-8736-1 fixing two Perl regex flaws that could cause denial of service, code execution, or security-restriction bypass.

Ubuntu released a security notice addressing two Perl vulnerabilities in regular expression handling. CVE-2026-15534 involves mishandling of large inputs during regex matching, enabling out-of-bounds heap reads or writes that could lead to denial of service or arbitrary code execution. CVE-2026-19487 involves incorrect matching for regexes with alternative branches, allowing security restrictions to be bypassed. Users are advised to apply the updated packages.

Ubuntu Security Notices·Sep 8, 14:25 UTC · 7d agoAdvisoryCVE-2026-15534CVE-2026-19487

CVE-2026-86219: Authen::SASL::Perl::DIGEST_MD5 versions before 2.2100 for Perl accept replayed authentication responses via unverified nonce in server_step

Authen::SASL::Perl::DIGEST_MD5 before 2.2100 for Perl accepts replayed DIGEST-MD5 authentication responses via unverified nonce handling (CVE-2026-86219).

CVE-2026-86219 affects Authen::SASL::Perl::DIGEST_MD5 versions before 2.2100 for Perl. The server_step function does not verify the nonce, allowing replayed authentication responses to be accepted in DIGEST-MD5 SASL exchanges. The fix is available in version 2.2100 of the perl-authen-sasl distribution.

oss-security·Sep 6, 18:55 UTC · 9d agoVulnerabilityCVE-2026-862191

CVE-2026-86287: Net::IP::LPM versions before 1.12 for Perl accept malformed prefix lengths

CVE-2026-86287: Net::IP::LPM Perl library before 1.12 accepts malformed IP prefix lengths, risking incorrect longest-prefix-match results.

Net::IP::LPM versions before 1.12 for Perl accept malformed prefix lengths, per an oss-security disclosure posted September 7, 2026. The lax input validation in this longest-prefix-match library could produce incorrect matching behavior. The fix is available in version 1.12 on CPAN.

oss-security·Sep 7, 18:40 UTC · 8d agoVulnerabilityCVE-2026-86287

CVE-2026-86304: MojoX::Authentication versions before 0.006 for Perl allow SAML authentication bypass because parse_assertion builds Net::SAML2::Binding::POST without a trust anchor

MojoX::Authentication before 0.006 for Perl allows SAML authentication bypass because parse_assertion builds Net::SAML2::Binding::POST without a trust anchor (CVE-2026-86304).

CVE-2026-86304 affects MojoX::Authentication versions before 0.006 for Perl. The parse_assertion function builds Net::SAML2::Binding::POST without a trust anchor, so SAML assertions are not validated against a trusted signing key, enabling authentication bypass. The flaw is fixed in version 0.006 of the module.

oss-security·Sep 6, 23:14 UTC · 9d agoVulnerabilityCVE-2026-86304

CVE-2026-16028: Protocol::HTTP2 versions before 1.14 for Perl allow memory exhaustion via closed streams that stream_state never removes from the connection stream table

CVE-2026-16028: Protocol::HTTP2 Perl library before 1.14 lets closed streams linger in the connection table, enabling remote memory exhaustion.

Protocol::HTTP2 versions before 1.14 for Perl allow memory exhaustion because closed streams are never removed from the connection stream table by stream_state. The flaw was disclosed on oss-security by Robert Rothenberg on September 7, 2026. Users should upgrade to Protocol-HTTP2 1.14 or later on CPAN.

oss-security·Sep 7, 18:45 UTC · 8d agoVulnerabilityCVE-2026-16028

Emacs arbitrary code execution: incomplete fix for CVE-2024-53920

Incomplete fix for Emacs CVE-2024-53920 allows arbitrary code execution when viewing or editing untrusted files in non-Lisp modes like Perl flymake.

Bas Alberts of GitHub Security Lab discovered that the patch for CVE-2024-53920, an arbitrary code execution flaw in Emacs, was incomplete. Code execution remains possible when viewing or editing untrusted text files in modes other than Emacs Lisp mode, for example Perl files with flymake mode enabled. All Emacs versions affected by the original CVE remain affected by the incomplete fix.

oss-security·Sep 14, 14:36 UTC · 1d agoVulnerabilityCVE-2024-53920

Eleethub: A Cryptocurrency Mining Botnet with Rootkit for Self

Unit 42 uncovered Eleethub, a Perl-based Bitcoin mining botnet using a crafted rootkit and libprocesshider to evade detection, still under development with few infections.

Unit 42 discovered an under-development Perl Shellbot campaign that mines Bitcoin using xmrig and emech while evading detection via a rootkit that replaces the ps tool and the LD_PRELOAD-based libprocesshider.so library. Infected devices download a malicious shell script, connect to IRC-based C2 at eleethub.com and the UnderNet network, and can receive commands for UDP/TCP floods, port scans, and HTTP attacks. The botnet can affect Unix systems and Windows 10 hosts running a Linux subsystem, and its operators use the 'Los Zetas' branding, though they are unlikely to be the Mexican drug cartel. Researchers found only a few compromised 'zombies' before the campaign expanded.

Palo Alto Unit 42·Aug 19, 12:05 UTC · 28d agoMalware in the wild

Debian 13.7 ships the fixes behind 92 security advisories, updates 106 packages

Debian 13.7 'trixie' point release bundles 92 security advisories and 106 package updates, including kernel, glibc, u-boot and qemu fixes.

Debian shipped version 13.7 of 'trixie', folding in 92 previously published security advisories and corrections to 106 source packages, including six Linux kernel advisories (DSA-6381, DSA-6393, DSA-6405, DSA-6415, DSA-6466, DSA-6477). glibc fixes a buffer overflow (CVE-2026-5928) and buffer underflow (CVE-2026-5450), with 17 packages rebuilt against the updated library; qemu carries 25 CVEs including a secure boot bypass (CVE-2026-16288), imagemagick 24, wolfssl 15 and perl 13. Boot-chain fixes include a u-boot FIT image verification bypass (CVE-2026-46728), a BOOTP/DHCP buffer overread (CVE-2024-42040), and corrected intermediate certificate verification in sbsigntool. The installer was rebuilt with kernel ABI 6.12.107+deb13, and existing systems receive the fixes through normal package mirror updates.

Help Net Security·Sep 14, 07:40 UTC · 2d agoAdvisoryCVE-2026-5928CVE-2026-5450CVE-2026-34873+17 CVEs

LexFlip: A Dissociation Diagnostic for Legal Meaning Preservation Metrics

LexFlip releases 373 minimal perturbations of Quebec statutory French that reverse legal force while preserving tokens, exposing weaknesses in embedding-based meaning preservation metrics.

LexFlip provides 373 minimal perturbations of Quebec statutory French that reverse legal force while preserving 0.93 of tokens, creating dissociation items that break monotone token-overlap metric validation. The seven embedding and BERTScore metrics tested register only 0.022-0.039 of their identical-to-unrelated range on these edits, versus 0.670 for bidirectional NLI. Against FrJudge, with a measured human ceiling of r=0.597, a bare length feature outscores every semantic metric tested.

arXiv cs.AI / cs.LG / cs.CL·Sep 4, 15:49 UTC · 11d agoAI research

1Password's AI patching benchmark is misleading

Trail of Bits reanalysis says 1Password's 26% AI clean-fix rate is misleading; 86% of eligible patches blocked exploits.

Trail of Bits critiques 1Password's FLAWED AI patching benchmark, arguing its 26% clean-fix headline mixes trials where agents were instructed to apply wrong fixes (22% of data) with trials that prohibited compiling or testing (36%). Restricting to reasonable conditions, 2,634 of 3,067 patches (86%) blocked the supplied exploit. Trail of Bits also reports 12.5% of 2,265 developer first fixes failed in its own 2024-2026 assessments, and released post-patch-validation and review-walkthrough agent skills.

Lobsters · security·Sep 15, 20:03 UTC · 17h agoResearch

OpenSSL’s new alpha build speeds up post-quantum crypto

OpenSSL 4.1.0 alpha1 adds DTLS 1.3 support and speeds up ML-DSA and ML-KEM post-quantum operations across several architectures.

The OpenSSL project released the first alpha of OpenSSL 4.1.0, months before general availability. The build adds DTLS 1.3 (RFC 9147) to secure UDP-based traffic such as VPNs, video calls and IoT, and includes optimized ML-DSA and ML-KEM operations for ppc64le, s390x and x86_64 plus AVX-512 acceleration for SHAKE hashing and AES-CBC decryption. Breaking changes include a new Net::Curl::Easy dependency for tsget, dropped Windows-on-Itanium and Windows CE support, and replacement of the no-ecdsa/no-ecdh Configure options with no-ec. It also adds GREASE support (RFC 8701), a DTLS mode for the SSL listener API, IKEv2 KDF support, and initial Elbrus2000 processor architecture support.

Help Net Security·updated Sep 10, 11:22 UTC · 6d ago·first Sep 9, 22:15 UTC · 6d agoTools 2 sources

Augustinian BabyLM: What Ostensive Definition Can and Cannot Teach a Small Language Model

Study shows visually grounded token embeddings in a small masked LM persist through training and improve object-property knowledge, but escape standard BabyLM benchmarks.

The paper implements ostensive definition for a small DeBERTa masked language model trained on 10M words, seeding visually grounded tokens with embeddings derived from labeled image regions before training. Visual initialization leaves a persistent, seed-replicated advantage on object-property knowledge (COMPS) and a corpus-tailored Visual-Property Swap benchmark covering color, material, size, and shape, but has no effect on most BabyLM grammar benchmarks. Synthetic grounding of previously unseeded words causally transfers the advantage to exactly those words.

arXiv cs.AI / cs.LG / cs.CL·Sep 10, 17:43 UTC · 5d agoAI research

ApateWeb: An Evasive Large-Scale Scareware and PUP Delivery Campaign

Unit 42 uncovers ApateWeb, a campaign using over 130,000 domains and multilayered redirects to deliver scareware, adware and PUPs to millions of users.

Unit 42 discovered ApateWeb, a large-scale campaign using a network of more than 130,000 domains to deliver scareware, potentially unwanted programs, adware including a rogue browser and browser extensions, and scam pages. The campaign uses a three-layer structure with deceptive emails as the entry point, centralized victim tracking via UUIDs, intermediate adware or anti-bot redirections, and evasion tactics such as cloaking, bot detection error pages, and wildcard DNS abuse. Activity spiked since August 2022, with several hundred attacker-controlled sites remaining in Tranco's top 1 million rankings and millions of monthly hits; Unit 42 blocked an estimated 3.5 million sessions across 74,711 devices in November 2023.

Palo Alto Unit 42·Aug 17, 12:48 UTC · Aug 17, 2026Phishing & fraud in the wild

Perplexity Details Its GPU Embedding Stack: How Ivy, Tulip and ROSE Serve pplx-embed

Perplexity details its GPU embedding serving stack (Ivy, Tulip, ROSE), which reuses LLM prefill/decode kernels, CUDA graphs, and LazyTensors to cut launch overhead.

Perplexity engineers published a deep dive on the serving infrastructure behind pplx-embed, used across Perplexity Search and its API platform. The stack comprises Ivy (Rust HTTP gateway), Tulip (gRPC scheduling and batching), and ROSE (Runtime-Optimized Serving Engine), which reuses LLM prefill and decode kernels rather than running a separate embedding engine. Optimizations include whole-model CUDA graphs with lazy capture and a LazyTensor abstraction that overlaps CPU batch preparation with in-flight GPU work. Benchmarks are reported against vLLM v0.22.0 in BF16, with FlashAttention 4 generally fastest but FlashInfer 3 winning on Qwen-based models at very long sequence lengths.

MarkTechPost·Sep 6, 03:20 UTC · 10d agoAI tools & infra1

Three in four AI-generated vulnerability patches leave something broken

Help Net Security·Aug 6, 00:00 UTC · Aug 6, 2026VulnerabilityCVE-2026-8512

Engineered Persuasion: Evaluating Personalized Pretexts in LLM-Generated Spear Phishing

A study of 180 US workers found each LLM phishing personalization level raised click-intention odds by 28%, but credibility depends on context fit.

The arXiv paper evaluates how personalized pretexts in LLM-generated spear phishing affect perceived credibility, using 180 US working adults across 1,436 evaluations of emails with four cumulative personalization levels, from workplace context to shared-project details. Convincingness rose 2.40 points per level in sensitivity analysis and click-intention odds increased 28% per level, while non-clickers shifted toward deleting rather than reporting. Qualitative coding showed details matching the recipient's role and routines supported credibility, whereas incorrect, vague, or channel-inappropriate details raised suspicion. The authors argue personalization effectiveness depends on pretext fit, with implications for workplace security training.

arXiv cs.CR·Sep 3, 19:21 UTC · 12d agoResearch

CiteShade: Citation Laundering in Multi-Source Retrieval-Augmented Generation and Its Counterfactual Defense

CiteShade attack makes RAG models cite trusted sources for attacker-chosen wrong answers, raising wrong-answer rate from 0.01 to 0.68.

CiteShade is presented as the first citation laundering attack against multi-source retrieval-augmented generation: an attacker controlling a single source induces a wrong answer falsely attributed to a trusted source, even while correct evidence remains in context. The attack is formalized via three necessary conditions (retrieval, generation, citation) constructible without any instructions, raising wrong-answer rate from 0.01 to 0.68 on multi-hop QA, with source deletion confirming the malicious source as causal driver. Vulnerability tracks a model's citation propensity rather than scale, reaching CLR 0.84 with explicit instruction and 0.64 without on the most citation-prone model. Perplexity filtering and citation-support checking prove insufficient; the authors propose a counterfactual defense verifying which source actually drove the answer.

arXiv cs.CR·Sep 14, 14:41 UTC · 1d agoAI safety & security1

Technical Manual for a Toolkit for Measuring Contextual Individuation in Transformer Language Models

An open methodology toolkit measures whether transformer language models contextualize fixed word forms across domains using bridge forms and layer-wise silhouette analysis.

The manual documents an open toolkit built around 'bridge forms' - identical written words recurring across two or more subject domains with a different sense in each - to test whether transformer language models individuate word occurrences by context beyond the embedding layer. It covers declarative specification of bridge forms, Wikipedia corpus acquisition, occurrence localization, layer-wise representation extraction, domain-pairwise silhouette measurement, and visualization, justifying each choice against failure modes such as sense contamination and subword-tokenization misalignment. It is a methodological and implementation reference and reports no empirical results.

arXiv cs.AI / cs.LG / cs.CL·Sep 4, 16:38 UTC · 11d agoAI research

Verifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering

Evaluation of twelve LLMs on 222 clinical questions shows verbatim quotes rarely substantiate claims; claude-opus-5 fully substantiates only 37.1%.

The authors build a standardized harness over four clinical practice guidelines and evaluate twelve LLMs on 222 synthetic clinical questions, measuring citation attachment, verbatim quote production, and claim substantiation. Most models attach verbatim quotes to over 90% of claims from prompting alone, though lightweight models like claude-haiku-4.5 struggle. Quotes frequently fail to substantiate claims: claude-opus-5 quotes 98.0% of claims but fully substantiates only 37.1%, exposing a capability gap for verifiable clinical QA.

arXiv cs.AI / cs.LG / cs.CL·Sep 14, 17:53 UTC · 1d agoAI research

PrivEscalate: Measuring and Augmenting the Threat of LLM-Automated Linux Privilege Escalation

Researchers release PrivEscalate, a 531-scenario benchmark showing LLM agents' Linux privilege-escalation success varies by vulnerability class, plus PrivEscAgent, a domain-specialized agent that boosts success.

The paper introduces PrivEscalate, an open-source benchmark of 531 Dockerized Linux privilege-escalation scenarios spanning 14 sub-categories, plus 329 parameterized variants measuring sensitivity to environmental distractors. Evaluating six LLMs across three agent architectures shows capability is heterogeneous across vulnerability classes, sensitive to perturbation, and architecture-dependent. The authors also present PrivEscAgent, a wrapper adding deterministic enumeration, category matching, and step planning that outperforms prior privesc-agent baselines without modifying the underlying LLM. The benchmark is released to support LLM agent evaluation, defensive tool validation, and red-team training.

arXiv cs.CR·Sep 8, 17:34 UTC · 7d agoResearch

I Am No One: Style-Aware Paraphrasing for Text Anonymization

Prompt-driven style-aware paraphrasing with LLMs cuts authorship attribution F1 by 60-70% while preserving content quality.

The paper proposes a style-aware, prompt-driven anonymization approach using pretrained LLMs to build compact stylistic profiles from minimal samples and rewrite text to suppress identifiable style markers while preserving meaning. It addresses stylometric re-identification risks in anonymized text, including ASR transcripts of meetings and call-center calls where leakage persists after acoustic anonymization. Across blog and review datasets, the approach reduces authorship attribution F1 by 60-70%, substantially outperforming both DP-based and non-DP baselines while maintaining readability.

arXiv cs.CR·Sep 11, 02:07 UTC · 5d agoResearch

MetroLLM-Bench: Evaluating Language Models as Transit Kiosk Runtimes

MetroLLM-Bench is a 955-case benchmark testing language models as transit kiosk tool-calling runtimes across six real metro systems.

The benchmark covers 37-414-station metro systems and eleven task categories including routing, fare calculation, disruptions, accessibility, and adversarial input, with 14 deterministic and 8 semantic scoring components. Of 26 models from six vendors, a PEFT-tuned 4B Qwen 3.5 student scored 91.3 on Tier 1, exceeding GPT-5.6 (90.6/90.0), while Muse Glimmer 30B led the composite ranking. A deterministic rule-based baseline reached 84.6, and PEFT gains over base models shrank from +7.03 points at 2B to -0.91 at 27B.

Hugging Face daily papers·Sep 8, 20:00 UTC · 7d agoAI research

Molecular Déjà Vu: Digit-Level Retrieval of Published Values in Frontier Language Models

Audit of 22 frontier models finds widespread verbatim retrieval of published molecular property values, with higher reasoning increasing recall of memorized numbers.

An arXiv audit tests 22 frontier LLMs across 12 molecular regression benchmarks for verbatim retrieval of published values. More than 50% of the LLMs show verbatim retrieval on five datasets, and identical experiments are flagged 89% more often at a high reasoning level than at the lowest one. Suppressing retrieval moves model prediction errors closer together in relative terms, suggesting predictive capability is not determined solely by memorized values.

arXiv cs.AI / cs.LG / cs.CL·Sep 4, 17:32 UTC · 11d agoAI research1

Right Tool, Right Job: Native-Language Evaluation, Tokenizer Sensitivity, and Methodological Findings from a French-Only BabyLM

French BabyLM entry METRON-FR (125M GPT-2, 92.47M words) shows tokenizer artifacts dominate child-scale zero-shot evaluation; proposes standard diagnostics.

METRON-FR is a 125M-parameter GPT-2 pretrained on 92.47M French words, submitted to the BabyLM 2026 Strict track, scoring 85.97% on the native Quebec-French QFrBLiMP benchmark and 62.80% on the BabyLM-weighted leaderboard. A cross-lingual GLUE protocol combining French task-data translation with rank-16 LoRA shows relational tasks gain while world-knowledge tasks regress. Bilingual Lexicon Induction reaches p@1 of 68.84%, 18x above chance, and ablations show single-token zero-shot scoring is dominated by tokenizer and template artifacts at child scale.

arXiv cs.AI / cs.LG / cs.CL·Sep 15, 16:43 UTC · 20h agoAI research

Quoting Rick Brewster

Paint.NET added a clean-room Direct2D rewrite for WINE, largely written by Anthropic's Claude and described as unreviewed 'vibe coded' code.

Rick Brewster says Paint.NET now ships a from-scratch, reverse-engineered Direct2D implementation (PaintDotNet.Windows.Direct2D1.Managed.dll) used under WINE via a /wine flag, since Direct2D was never completed well enough there. He credits the Claude coding assistant with writing most of the code, calling it largely 'vibe coded' and not thoroughly reviewed. Simon Willison shared the quote as an example of shipping AI-assisted systems code in production software.

Simon Willison·Sep 2, 05:50 UTC · 14d agoAI tools & infra1

An Empirical Analysis of CodeQL False Positives and Query Refinements for Java Vulnerabilities

Study of 167 Java CVE instances finds CodeQL false positives follow recurring patterns; query refinements remove 81.8% of reviewed ones.

Researchers ran CodeQL's Java security query suite on 167 CVE instances from 110 projects, manually reviewing 500 sampled false-positive paths and building a five-category taxonomy led by Missed Path Constraint or Sanitization (36.6%), Benign Execution Context (29.4%), and Missing Trust Boundary Modeling (27.6%). Guided by the taxonomy, query-level refinements removed 81.8% of reviewed false positives and 15.8% of reported paths across the selected queries while retaining 7 of 8 true positives. To address generalization, agentic coding tools given the refinement patterns as templates adapted them to new projects successfully in 56% and 62% of tasks, versus 28% without guidance.

arXiv cs.CR·Sep 3, 22:48 UTC · 12d agoResearch1

You've Got a BUD in Me: Authenticated Reads from Per-Block Write Logs

Researchers propose BUD, per-block write-log digests enabling blockchain validators to serve historical membership and exclusion proofs far cheaper than state-wide tries.

The paper introduces Block Update Digests (BUD), which authenticate each block's write log with predecessor pointers, plus a SuperBUD and exponential hierarchy to turn long unchanged intervals into short proofs. Soundness against adversarial provers and up to f Byzantine validators is proven under archive, attestation, and committee evidence assumptions. Benchmarks show a 50x state-size increase raises the base-BUD path only 1.24x versus 3.1x for in-memory and 69.5x for disk-backed Merkle Patricia tries, with read payloads below 800 bytes and p99 warm verification at 146 microseconds.

arXiv cs.CR·Sep 10, 08:46 UTC · 6d agoResearch

PARSER: Read in Parallel, Reason in Depth for Long-Context LLM Agents

PARSER uses parallel reader subagents and an RL-trained lead agent for long-context QA, beating baselines and cutting latency up to 11x.

The PARSER paper decouples reading from reasoning: frozen subagents each read one document chunk in parallel while an RL-optimized lead agent iteratively broadcasts queries and aggregates evidence in scatter-gather rounds. On multi-hop QA with 7K to 896K token contexts, a 4B-backbone PARSER beats the strongest sequential memory baseline by 5.7 points on average and 12.0 points at 896K tokens, and a 9B version surpasses DeepSeek-V4-Pro by 6.3 points. Controlled experiments show robustness to evidence position, order, and distance perturbations, with inference latency reduced by up to 11x.

Hugging Face daily papers·Sep 5, 20:00 UTC · 10d agoAI research1

Everything in Moderation: Per-Domain Coverage Optima and Alignment-Resistant Domain Gaps in Multi-Domain Mid-Training

Controlled mid-training experiments on Qwen3-8B-Base find each domain has a 10-40% coverage optimum and domain gaps survive alignment SFT.

Using Qwen3-8B-Base (with a 4B replication) across five semantically rule-disjoint KOR-Bench domains, the authors train 30 data allocations spanning the five-domain simplex at five seeds each. All five domains show interior optima in the moderate 10-40% coverage band, and domain gaps persist after a fixed-budget compensatory SFT pass, which raises 116/120 cells yet bridges 0/240 pairs at a 5% threshold. Zero coverage collapses mid-training-only accuracy, though a FineWeb-Edu-only control shows the collapse is partly generic drift. The results argue mid-training data composition requires principled design rather than reliance on later alignment.

arXiv cs.AI / cs.LG / cs.CL·Sep 8, 17:31 UTC · 7d agoAI research1

Access Control as Verified Parse Constraints

Researchers verify a class of EverParse validators that correctly enforce access-control policies, deploying a machine-checked enforcement gate on seL4.

The paper targets enforcement-code bugs in commercial security gateways by proving that forward-only, backtrack-free EverParse validators are verified recognizers for a bounded finite-state class that includes access-control decision functions with fixed-offset fields and bounded disjunction. Encoding a bounded policy language into a fixed-size byte buffer allows an SMT solver to verify the enforcement code once, covering all byte values, policies, requests, and sessions. Editing rule content over a fixed endpoint set requires no new proof, while adding endpoints reruns the toolchain. A deployment on the seL4 microkernel ensures every request passes through the gate and unverified components cannot corrupt the enforcement chain.

arXiv cs.CR·Sep 11, 06:33 UTC · 5d agoResearch

LLM Agents as Computational Typologists

AUTOTYPOLOGIST is an LLM agent that performs evidence-grounded linguistic typology analysis over 25 open-source reference grammars.

The agent retrieves relevant grammar sections, analyzes interlinear glossed text (IGT), and iteratively reasons over typological hypotheses in a ReAct-style workflow. It was evaluated on typological feature coding against expert annotations and hypothesis testing against universals using 25 open-source reference grammars. Results suggest LLM agents can support scalable, inspectable crosslinguistic analysis but still require expert validation.

arXiv cs.AI / cs.LG / cs.CL·Sep 7, 17:35 UTC · 8d agoAI research1